On Tuesday, the president of Opeli, Altman himself announced a number of novel user rules, including a commitment to a significant change in the method of CHATGPT interaction with users under 18 years of age.
“We prioritize the safety of teenagers’ privacy and freedom,” we read in post. “This is a new and powerful technology and we think that minors require significant protection.”
Changes in juvenile users concern in particular conversations on sexual topics or self -mutilation. According to the novel CHATGPT policy, he will be trained so as not to engage in a “flirtatious conversation” with juvenile users, and additional handrails will be placed around the discussion about suicide. If a juvenile user uses chatgpt to imagine suicidal scenarios, the service will try to contact parents or, in particularly harsh cases, local police.
Unfortunately, these scenarios are not hypothetical. Opeli is currently in the face of the unlawful death of the death of Adam Raine parents, who died due to suicide after months of interaction with chatgpt. The character. Ai, another consumer chatbot, is in the face of a similar process. While the risk is particularly urgent for juvenile users considering self -mutilation, the wider phenomenon of Chatbot’s illusion caused widespread concern, especially when consumer hut became capable of more sturdy and detailed interactions.
Along with restrictions based on the content, parents who register the juvenile user account will have the power to set the “darkening hours” in which chatgpt is not available, a function that was not previously available.
Recent CHATGPT rules appear on the same day Hearing of the Senate of the Judiciary entitled “AI Chatbot Damage Examination”. The trial was announced by Senator Josh Hawley (R-MO) in August. Adam Raine’s father is to speak at the trial, including guests.
The trial will also focus on the arrangements Reuters investigation These discovered political documents apparently encourage sex conversations with juvenile users. The finish has updated its chatbot rules after the report.
TechCrunch event
San Francisco
|.
October 27-29 2025
Separation of underage users will be a significant technical challenge, and Openai described its approach in detail In a separate post on the blog. The service is “built towards a long -term system to understand if someone has exceeded 18 years”, but in many ambiguous cases the system will be default in relation to more restrictive rules.
For those interested, the most reliable way to make sure that a juvenile user will be considered a combination of a teenager account with an existing parental account. This also allows the system to directly warn parents when a teenage user is believed to be in danger.
But in the same post, Altman emphasized OpenAi’s constant involvement in the privacy of users and giving adult users wide freedom in interaction with chatgpt. “We realize that these rules are in conflict,” sums up the post, “and not everyone will agree with how we solve this conflict.”